Prediction Of Temporal And Spatial Distribution Of Passenger Flow Under The Condition Of Urban Rail Transit Interruption

As an important part of the urban comprehensive transportation system,the travel paths of a large number of passengers will be destroyed in the event of an operation interruption accident,which will cause changes in the distribution of passenger flow in the network,resulting in stagnation,congestion,and even serious safety accidents.In order to ensure the safety of passengers,the emergency management and operation departments will take emergency measures as soon as possible.Therefore,accurately predicting the temporal and spatial distribution of passenger flow after the interruption will help relevant departments to formulate emergency plans scientifically and rationally,and adjust treatment measures in a targeted manner.To this end,this paper takes the urban rail transit section operation interruption as the research background,and carries out the following aspects of work on the prediction of passenger flow spatiotemporal distribution:This paper firstly constructs a short-term OD demand forecasting model under the condition of urban rail transit interruption.Based on the theory of tensor CP decomposition and completion,the model introduces the site weak correlation penalty term and graph Laplacian penalty for optimization;a model solving algorithm based on block coordinate descent is designed;the grid search method is used for parameter optimization.Experiments show that the model has applicability and effectiveness in solving the problem of OD demand forecasting under interruption conditions,and provides a short time granularity and good forecasting accuracy.Secondly,by analyzing the emergency management measures under the condition of interval interruption,the paper summarizes 6 travel plans of the affected passenger flow during the interruption period,and constructs the space-time expansion network of interval interruption;redefines the effective path constraints under normal and interruption conditions;based on The K-shortest path finding algorithm searches for efficient paths and space-time paths.Then,based on the nested Logit model,a travel plan selection model for the affected passenger flow is constructed,and a prediction algorithm for the spatial-temporal distribution of passenger flow under the condition of interval interruption is designed.In this paper,a travel route selection tree consisting of 3 strategies and 6 plans is established;the parameters are calibrated using SP survey data and STATA software;according to the calibration results,the influencing factors and decision-making mechanism of the travel plan selection for the affected passenger flow are analyzed;the results show that the nested Logit model is more suitable for studying the problem of travel route selection during interruptions.In addition,a "normal-interruption-recovery" three-stage passenger flow redistribution algorithm is designed to predict the spatiotemporal distribution,and output the passenger flow distribution index.Finally,taking Shanghai urban rail transit as the background,the section interruption event is simulated,and the spatiotemporal distribution of passenger flow under normal and interruption conditions is predicted.The forecast results of section passenger flow,transfer passenger flow,inbound and outbound passenger flow and other indicators are compared and analyzed,and the law of changes in the spatial and temporal distribution of passenger flow under the condition of interval interruption is summarized;it provides reference for emergency response measures for the interruption event and puts forward targeted suggestions.